Preserve a workspace archive and recover its identity and knowledge.
A backup captures the workspace material needed for recovery under its declared scope. Keep the archive and required key material available to authorized recovery operators.
A live replica follows ongoing changes. A preserved backup gives you a separate recovery point.
Replication policies help services survive covered failures while repairing copy deficits. Archives preserve an independent restoration path when continuity is insufficient.
Recovery restores the workspace under its existing identity. Check that the archive is complete, that decryption authority is available, and that the recovered state passes integrity checks before using it.
For example, recover a project workspace after losing its serving machines, then inspect the recovered history and membership.
Copying selected data into a new workspace creates a different workspace. Retention and erasure determine what must remain or be removed. Neither is the same operation as restoring an archive.
A recovery plan must cover the archive's location, permissions, keys, and verification. High availability covers continuity while services fail.
Developer reference
The archive contains original encrypted workspace blocks plus durable interpretation metadata. Workspace identity, committed branch heads and history survive restoration. This is physical recovery, not RDF export/reimport into a newly identified workspace.
Restore verifies completeness and content-addressed integrity, then reinstates preserved state without replaying business operations or replacing original authorship/timestamps. Reading recovered data still requires authorized key material.
A cold archive occupies a separate failure domain from live replication. A deletion propagated among online replicas need not remove independently retained archive blocks. Define archive access, retention and disposal explicitly, and exercise restoration through usable queries rather than treating a successful upload as a recovery test.